NOAA National Weather Service (NWS) field offices produce estimates of gridded hourly observed precipitation by objectively combining rain gage data with radar and satellite-based data to create a set of near-real-time multi-sensor estimates. These grids are subject to extensive interactive quality control operations which ultimately produce the “best estimate” grid for a given hour. The final quantitative precipitation estimate (QPE) grid is generated primarily for use in river forecast models and is also made available to the public via official products and the Internet.

This paper discusses many important additions to the QPE generation process implemented by the NWS Office of Hydrologic Development (OHD) over the last few years. These additions include incorporation of locally-developed analysis techniques and interactive tools into a unified Multi-sensor Precipitation Estimator (MPE) application for national use. Interactive tools are used to apply quality control checks and for adjusting the resulting quantitative estimate. The paper also describes the steps for generating the official best estimate QPE product which is then distributed.

Secondary topics also discussed include recent operational advances in the use of bias adjustments to improve data quality, methods for time disaggregation of station precipitation reports, and the expanded ability of MPE to provide temperature and freezing level information beyond its core precipitation focus. Other topics briefly covered are the relationship of this hourly product to other NWS precipitation products with varying spatial and temporal resolution, the use of externally provided products for use as the initial QPE estimate, and expanded interaction with satellite-based estimates. Lastly, planned areas of future work are described, including use of gridded QPE in a new NWS river modeling system which makes use of alternate spatial resolutions and data formats.